Systematic detection of Titan's clouds in VIMS/Cassini hyperspectral images using a new automated algorithm

Abstract

Titan is the Saturn's largest moon where meteorological processes are very active, as observed most recently by the Cassini/Huygens orbiter. Cloud monitoring is a prime method to observe, describe and understand present climate on Titan. Unlike our previous detection method, which was based on manual control of threshold, we investigate here the possibility of a fully automated methodology based on blind source separation to analyzing years of Cassini near-infrared cloud images. Since the spectral signature of Titan clouds are diverse and not known a priori, the choice of a blind source separation seems to be appropriate. Preliminary results show that Titan's cloud detection is possible using the recent implementation of a Bayesian source separation method.